Outer Step 0. Inner Step 154. Batch Size 96

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5156091,
-    5156102,
-    5156108,
-    5156114,
-    5156119,
-    5156128,
-    5156136,
-    5156149,
-    5156156,
-    5156163
+    5156190,
+    5156198,
+    5156203,
+    5156208,
+    5156213,
+    5156218,
+    5156222,
+    5156228,
+    5156233,
+    5156237
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 147,
+  "inner_step": 154,
   "inner_steps": 0,
   "last_allreduce_block": 4016775,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:05f91c2464234842d488dae508c119226160fa5549e1c485ebdf6e78e38b25a9
+oid sha256:d6de57a417037a20092a6f4fcd602ec8f59b1c493760bbd4fdcac54e4dbc6709
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a22cd342d8295fe83d6ac66e3b8383a934fd2cbb3f16db1824bc9899ed8f754b
+oid sha256:a57292481ecdd1d6445f4d87b9f799ddb94af7191beb0e7f6216dbd62b78a8ee
 size 4040701744